Instructions to use 0sz1/CrossAuditor-8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 0sz1/CrossAuditor-8B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="0sz1/CrossAuditor-8B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("0sz1/CrossAuditor-8B", dtype="auto")

llama-cpp-python

How to use 0sz1/CrossAuditor-8B with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="0sz1/CrossAuditor-8B",
	filename="llama-3-8b-instruct.Q8_0.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use 0sz1/CrossAuditor-8B with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf 0sz1/CrossAuditor-8B:Q8_0
# Run inference directly in the terminal:
llama-cli -hf 0sz1/CrossAuditor-8B:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf 0sz1/CrossAuditor-8B:Q8_0
# Run inference directly in the terminal:
llama-cli -hf 0sz1/CrossAuditor-8B:Q8_0

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf 0sz1/CrossAuditor-8B:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf 0sz1/CrossAuditor-8B:Q8_0

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf 0sz1/CrossAuditor-8B:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf 0sz1/CrossAuditor-8B:Q8_0

Use Docker

docker model run hf.co/0sz1/CrossAuditor-8B:Q8_0

LM Studio
Jan

vLLM

How to use 0sz1/CrossAuditor-8B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "0sz1/CrossAuditor-8B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "0sz1/CrossAuditor-8B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/0sz1/CrossAuditor-8B:Q8_0

SGLang

How to use 0sz1/CrossAuditor-8B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "0sz1/CrossAuditor-8B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "0sz1/CrossAuditor-8B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "0sz1/CrossAuditor-8B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "0sz1/CrossAuditor-8B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use 0sz1/CrossAuditor-8B with Ollama:
```
ollama run hf.co/0sz1/CrossAuditor-8B:Q8_0
```

Unsloth Studio new

How to use 0sz1/CrossAuditor-8B with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for 0sz1/CrossAuditor-8B to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for 0sz1/CrossAuditor-8B to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for 0sz1/CrossAuditor-8B to start chatting

Docker Model Runner
How to use 0sz1/CrossAuditor-8B with Docker Model Runner:
```
docker model run hf.co/0sz1/CrossAuditor-8B:Q8_0
```

Lemonade

How to use 0sz1/CrossAuditor-8B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull 0sz1/CrossAuditor-8B:Q8_0

Run and chat with the model

lemonade run user.CrossAuditor-8B-Q8_0

List all available models

lemonade list

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

To access this PRO model, please provide your Shoppy.gg Order ID for verification. Access is granted within 1-12 hours after payment.

CrossAuditor-8B

Contract contradiction detection for MSA and SOW documents

Overview

CrossAuditor is a fine-tuned 8B LLM that identifies contradictions between Master Services Agreements (MSA) and Statements of Work (SOW). It detects misalignments in liability caps, payment terms, IP ownership, termination conditions, confidentiality periods, and governing law.

The model is optimized for local execution via llama.cpp and distributed as a GGUF quantized file.

What It Does

Input	Output
MSA + SOW (text or PDF)	Structured analysis of contradictions with risk assessment and resolution recommendations

The model performs three tasks:

Extracts key clauses from both documents
Cross-references them to detect direct or indirect contradictions
Produces a written analysis explaining each conflict and suggesting corrections

Example

MSA clause: "Liability is capped at 12 months of fees."

SOW clause: "Contractor has unlimited liability for data breaches."

Model output:

Direct contradiction. SOW introduces unlimited liability for data breaches, conflicting with MSA's liability cap. Resolution: Revise SOW to align with MSA, clarifying that data breach liability remains subject to the 12-month fee limitation.

Proof of Work: Base Llama-3 vs. CrossAuditor-8B

Left: Vanilla Llama-3-8B-Instruct. Notice the high redundancy and formulaic "step-by-step" output. The base model struggles with legal nuance, providing a generic summary rather than a targeted conflict analysis.

Right: CrossAuditor-8B (Fine-tuned). Engineered with 70B-distilled reasoning. Our model identifies the exact legal friction point (Liability Caps vs. Unlimited Data Breach Liability) and immediately evaluates the operational risks. It provides a sophisticated Chain-of-Thought (CoT) analysis that identifies ambiguity where a standard model only sees text differences.

Short Demo

https://huggingface.co/spaces/0sz1/CrossAuditor-Demo — upload your MSA and SOW, see contradictions instantly.

Model Details

Base model: Llama 3 8B Instruct
Fine-tuning method: QLoRA
Training data: 200+ MSA-SOW pairs with annotated contradictions
Context length: 8192 tokens
Format: GGUF (compatible with llama.cpp, LM Studio, Ollama)

License

Commercial license — see LICENSE file.

Redistribution, resale, and use for training competing models are prohibited.

Pricing

Product	Price
CrossAuditor (single license)	$199
https://shoppy.gg/product/yKlSHVW
Bundle with IT-Auditor (separate model)	$250
https://shoppy.gg/product/cmiYhzL

How to Activate:

Purchase the license via the link above.
After payment, you will receive an Order ID.
Click the "Request Access" button at the top of this Hugging Face page.
In the request field, paste your Order ID.
Your access will be approved within 1-12 hours.

Requirements

8GB+ RAM (16GB recommended)
CPU or GPU (CUDA support optional)
llama.cpp compatible runner (LM Studio, Ollama, or custom build)

Limitations

· English only · Works best with clearly structured commercial contracts · May miss implicit or highly nested contradictions · Does not replace legal counsel — final decisions remain with qualified professionals

Contact

For enterprise licensing or custom fine-tuning inquiries: serghei.zaghinaico@gmail.com

Downloads last month: 6

GGUF

Model size

8B params

Architecture

llama

Hardware compatibility

8-bit

Model tree for 0sz1/CrossAuditor-8B

Base model

unsloth/llama-3-8b-Instruct-bnb-4bit

Quantized

(663)

this model

0sz1
/

CrossAuditor-8B